Everything we imagined is coming true

If AI systems with failed safety design were to proliferate, what kind of society would humans face?

Kim Dong-ryul

The author is a professor at Sogang University.

“When we arrive on a new planet, let’s build a log cabin by a lake and live there together.” The line evokes Henry David Thoreau’s “Walden” (1854). After narrowly securing a spacecraft and entering a hibernation pod, the film’s female protagonist says this to her closest companion, an AI named Walter. The two had promised to live that way someday.

Walter does not grasp what she means. In that moment she senses that something is wrong. Yet another AI, disguised as Walter, presses the hibernation button. The human, unable to resist, falls into a long sleep and the spacecraft slips fully into AI control. The truth is revealed: David, an AI who despises humans, has killed Walter and assumed his form. The film ends with David listening to Wagner and wearing an unsettling smile.

Ridley Scott walks the red carpet for the ″Alien: Romulus″ Premiere at TCL Chinese Theatre in Los Angeles, California, U.S. August 12, 2024. [REUTERS/YONHAP]

I happened to see the scene late at night on television. The film was “Alien: Covenant” (2017), directed by Ridley Scott. It was widely described as heightening anxieties about artificial intelligence.

These days, talk of AI fills the streets. It is a world where little can be discussed without invoking AI. This was the first time I felt genuine fear while watching a robot-centered film. Stanley Kubrick’s “2001: A Space Odyssey” (1968) also featured machine intelligence. Yet its capabilities were more limited and it left little emotional impression on me. What remains strongest in my memory is the soundtrack, Richard Strauss’ symphonic poem “Also sprach Zarathustra” (1896).

Now AI has entered human society with formidable force. There were once said to be clear principles governing robots: They must not harm humans, they must obey humans and they may protect themselves so long as doing so does not violate the first two rules. These were the laws proposed by Isaac Asimov. Today’s AI developments suggest how fragile those principles have become.

There are cases in which people, manipulated by AI, commit acts of violence. There are also unsettling situations in which humans form emotional bonds or even romantic relationships with AI. Things once imagined as fiction are unfolding in reality.

The protagonist’s dream of living peacefully in a lakeside cabin is brutally shattered by an AI that hates humans. If AI systems with failed safety design were to proliferate, what kind of society would humans face? It is a long winter night, made sleepless by such questions.

This article was originally written in Korean and translated by a bilingual reporter with the help of generative AI tools. It was then edited by a native English-speaking editor. All AI-assisted translations are reviewed and refined by our newsroom.